You've raised capital. Your device works. Your team is ready to launch. There's one gatekeeping step left: FDA clearance, De Novo classification or approval. If you've never navigated FDA before, it can look straightforward: submit documentation, FDA reviews, and 90 days later you're cleared to sell. In practice, first submissions often draw requests for additional information, new data demands and resubmissions that stretch the timeline and the budget at the moment a company needs both.
This article is written for founders. It covers five common FDA submission mistakes, what each tends to cost in time, and how an early Pre-Submission can surface them. The examples are illustrative composites, not specific companies.
Five Common FDA Submission Mistakes
Mistake #1: Choosing the Wrong Regulatory Pathway
Many founders think there are only two FDA pathways: 510(k) (fast) and PMA (slow). The three main premarket routes are 510(k), De Novo and PMA, and picking the wrong one is expensive.
The mistake: You build a novel device with no clear predicate device. To avoid complexity, you force-fit a 510(k) submission with a weak or inappropriate predicate. FDA reviews the submission and finds the device not substantially equivalent to the predicate. You're pushed into a De Novo request or a PMA, adding months and cost.
Illustrative example: A diagnostic startup submitted a 510(k) for a novel AI-powered pathology analysis device, claiming substantial equivalence to a manual pathology review system. In FDA's view, the AI analysis had no valid predicate, so the comparison failed. The company was pushed into a De Novo request, adding months to its timeline and requiring clinical validation it hadn't planned for.
Right Pathway Selection Saves Time and Money
FDA user fees for fiscal year 2027 (October 1, 2026 through September 30, 2027), standard rate: 510(k) $28,653 · De Novo $191,020 · PMA $636,732. Qualifying small businesses pay 25% of the standard fee. FDA's review goals are 90 FDA days for a 510(k), 150 for a De Novo (which creates a new classification) and 180 for a PMA; total time runs longer when a request for additional information puts the review on hold.
Mistake #2: Picking an Inappropriate Predicate Device
For 510(k) submissions, you must identify a legally marketed predicate device that your device is substantially equivalent to. "Substantially equivalent" doesn't mean clinically superior. It means the device has the same intended use as the predicate and either the same technological characteristics or different ones that don't raise different questions of safety and effectiveness. This is where founders commonly stumble.
The mistake: You pick a predicate that's similar in function but not in technology. FDA questions whether your device's novel mechanism of action is substantially equivalent to the predicate's mechanism. They request additional clinical data to demonstrate equivalence. A 510(k) planned around a 90-day review stretches out over rounds of information requests.
Illustrative example: A spinal fusion device startup selected a predicate device that was also a spinal fusion system, but used a completely different fixation mechanism. The predicate used screw-and-rod fixation; the new device used interlocking bone implant technology. While both addressed the same medical indication, the technological divergence was too great. FDA issued an information request requiring biomechanical studies, animal testing, and clinical data to demonstrate equivalence. The timeline stretched from a planned 90-day review to more than a year.
Predicate Selection Red Flags
A predicate removed from the market at FDA's initiative, or found adulterated or misbranded by a court, can't support a 510(k). FDA's draft predicate-selection guidance also looks for predicates cleared with well-established methods, meeting expected safety and performance, free of unmitigated use-related or design-related safety issues, and without a design-related recall. A predicate with a very different mechanism of action invites questions too.
Mistake #3: Underestimating Clinical Data Requirements
Founders often assume that if their device is 510(k)-eligible, they need minimal clinical data—maybe some published literature and a small bench study. FDA may disagree. The statutory standard for 510(k) is "substantial equivalence," and the evidence FDA expects depends on the device and how it differs from its predicate.
The mistake: You submit a 510(k) with literature review and bench testing, assuming that's sufficient. FDA's information request: "Provide clinical data comparing your device to the predicate in [specific patient population]." You don't have that data, and generating it takes months.
Illustrative example: A glucose monitoring startup submitted a 510(k) with bench accuracy data and reference to published literature on similar devices. FDA's information request asked for a clinical study comparing the device's accuracy with the predicate's in patients with diabetes, across glycemic ranges. The company hadn't planned for this. Designing, running and analyzing a clinical study added months and significant cost.
Prevention: before you submit, request a Pre-Submission. There is no FDA user fee for it, and FDA gives written feedback, plus a meeting if you ask for one, on the approach you describe: whether the clinical data plan looks sufficient, whether the predicate is appropriate, or whether a De Novo request looks more suitable.
Mistake #4: Missing Biocompatibility Testing Requirements (ISO 10993)
For any device that contacts patient tissue (skin, blood, bone, etc.), you must demonstrate biocompatibility under ISO 10993 standards. This isn't optional. Yet many founders build this testing into their timeline only after FDA raises it.
The mistake: You submit your FDA application without mentioning biocompatibility testing. FDA's information request: "Provide ISO 10993 biocompatibility testing reports for all materials contacting patient tissue." You assumed your materials were "safe" (they probably are), but biocompatibility needs a documented evaluation under ISO 10993-1, usually with testing at a qualified laboratory. You're now months behind schedule.
Illustrative example: A breast reconstruction implant startup submitted a PMA without comprehensive ISO 10993 testing for the new silicone elastomer used in their device. FDA asked for testing that included cytotoxicity (ISO 10993-5), sensitization (ISO 10993-10), systemic toxicity (ISO 10993-11) and irritation (ISO 10993-23). The company thought "tested for biocompatibility" meant they were done. Testing and reporting added months to the timeline.
Mistake #5: Not Understanding "Substantial Equivalence"
This is the most conceptually misunderstood standard. Substantial equivalence doesn't mean "as good as" or "similar to." It's a regulatory standard: the intended use must be the same, and the technological characteristics must be the same or, if different, must not raise different questions of safety and effectiveness.
The mistake: You believe your device is substantially equivalent because it solves the same clinical problem as the predicate. But if your technology is fundamentally different, substantial equivalence collapses. FDA finds it not substantially equivalent, and you're in De Novo or PMA territory.
Illustrative example: A continuous glucose monitoring device company submitted a 510(k), claiming substantial equivalence to a finger-stick blood glucose meter predicate. Same intended use (glucose monitoring), but radically different technology (subcutaneous sensor vs. capillary blood sample). FDA found the devices not substantially equivalent, pointing the company to a De Novo request. The company had assumed "same clinical use = substantial equivalence." It doesn't work that way.
How Founders Can Avoid These Mistakes
The solution is surprisingly simple: use FDA's Pre-Submission process, part of the Q-Submission program. There is no FDA user fee for it. Before spending resources on your formal submission, you present:
- Your proposed regulatory pathway (510(k), De Novo, or PMA)
- Your predicate device (if 510(k)), with justification for why it's appropriate
- Your clinical data plan
- Your biocompatibility testing plan
- Your substantial equivalence arguments
FDA responds with written feedback, which can say the proposed pathway looks appropriate, that a De Novo request looks more suitable, or that the clinical data plan needs more. Requesting it well before a planned submission leaves time to adjust.
An Illustrative Case: Using a Pre-Submission Early
A neuromodulation startup requested a Pre-Submission well before its planned submission. FDA's feedback flagged the 510(k) approach as risky given the device's novel waveform and pointed toward a De Novo request, which creates a new classification. The company adjusted, submitted a De Novo with the supporting data FDA had asked about, and the request was granted. Had it gone ahead with the 510(k), it would likely have faced a not-substantially-equivalent decision and a second submission.
The Bottom Line for Founders
FDA submissions aren't mysterious, but they're unforgiving. A mistake in pathway selection or predicate choice can cascade into months of delay and unplanned cost. A Pre-Submission carries no FDA user fee and takes weeks of preparation; recovering from a poor initial choice can take far longer.
Your device probably doesn't change because of FDA feedback. But your timeline, your budget, and your pathway almost certainly will. Plan for it upfront.
"The best time to learn FDA requirements is before you submit, not after FDA finds a problem with your application."
References
- FDA. "Premarket Notification 510(k)." fda.gov
- FDA. "De Novo Classification Request." fda.gov
- FDA. "Premarket Approval (PMA)." fda.gov
- FDA. "Use of International Standard ISO 10993-1, 'Biological evaluation of medical devices - Part 1: Evaluation and testing within a risk management process'." Guidance, September 2023. fda.gov
- FDA. "Requests for Feedback and Meetings for Medical Device Submissions: The Q-Submission Program." Final guidance, May 2025. fda.gov
- FDA. "Medical Device User Fee Amendments (MDUFA) Fees." fda.gov
- Federal Register. "Medical Device User Fee Rates for Fiscal Year 2027." July 30, 2026. federalregister.gov
Need a Custom Analysis?
Vantage full reports assemble the public record on a company, cite factual claims to their sources, and are reviewed by Arvind Ravinutala, MD, and Aswini Ravinutala, BME. Start with a free 20-minute call.